Switching between 2 resistors using P channel FET

[sean87]

Hi all,

I am dealing with kind of a rough situation. I have a power opamp that outputs a 50V 4A DC signal. At the output stage of the opamp I need to select between a 1k resistor and a 10k resistor right before the load, something like this:



I mean I want either the output stage of opamp and load be in series with the 1k resistor or the 10k resistor. I tried to simulate this with P-Channel MOSFET but any combination of opening/closing switches does not have any effects! Can you please help me with designing this circuit? I need to open/close the MOSFETs using an Arduino, something like this?

[danadak]

You are high side switching, so need adequate gate drive to turn
on the switch. Application note -


http://www.vishay.com/docs/70611/70611.pdf


http://www.onsemi.com/pub/Collateral/AND9093-D.PDF


Regards, Dana.

[eblc1388]

The following circuit will work only when vcc is positive and higher than 20V.

PMOS M1 & M2 and transistor Q1 & Q2 should have voltage rating higher than 90V, and current rating of PMOS rated current higher than your requirement.

D1 & D2 is 12V or 15V Zener diode for gate overvoltage protection.

X1 & X2 represent 5V control signal coming from the Arduino port pins.

[BrianHG]

BTW, you only need 1 mosfet unless you want 3 switch options. 1k load, 10k load, and open.  Otherwise, hard-wire the 10k always on, and replace the 1k with either 1.11k.  This saves you a mosfet and a gate driver circuit.

As for an alternate AC switch, one which wont need careful proper tracking gate driving, you can try a mosfet output optocoupler like one of these:
https://www.digikey.ca/product-detail/en/toshiba-semiconductor-and-storage/TLP222AF/TLP222AF-ND/871243

[T3sl4co1l]

How do you have a "50V 4A" signal if it's going through minimum 0.91kohms (= 55mA worst case)?

Bandwidth is unstated, so it's anyone's guess if this is DC (a mechanical or MOS relay would do) or RF (mechanical relay, or a much more complicated active circuit).

Switching speed is also unstated.

Polarity is also unstated, though if your diagram is to be taken at face value, it seems the op-amp only outputs negative voltages?  (In other words, your circuit won't "switch" while the op-amp output is positive: that's what the little triangles in the MOSFET symbol are there for, they indicate the internal body diode.)

If it is bipolar (positive and negative voltage wrt ground), you need a bidirectional switch.  Usually, a pair of MOSFETs are connected together source-to-source, with the switch terminals being the two drains.  An isolated gate driver is needed to turn it on and off.  PhotoMOS SSRs do this, with a handy optocoupler, all in one package.  But they're quite slow (~milliseconds), so they're not suitable for everything.

If you only need two resistor settings (not open circuit), consider using one "default" resistor, and switching another in parallel with it.  To get 1 or 10k, you'd use 10k and 1.11k resistors, with only one switch for the 1.11k resistor.

You may be better off harnessing features of the power op-amp.  If it has a "mute" or "disable" or "power down" option, that should leave the output open-circuit, which is the same as opening a switch.  Use one amp to drive each resistor you need.  This has the advantage that it's fast (as fast as the internal circuitry), and is limited only by op-amp performance (so a similar method can be used all the way up into RF, if needed).

Tim

[Brutte]

I have a power opamp that outputs a 50V 4A DC signal. At the output stage of the opamp I need to select between a 1k resistor and a 10k resistor right before the load, something like this:

I really do not get the idea.
For pushing in 4A at 50V you need the load that is 12R5. So looks like with R3 of positive resistance that is impossible in your scheme. Or is R3 negative?
This op-amp controls the load through drop resistor/s. Why is that so? Are these shunt resistors for current sensing but the feedback has not been shown on a schematics?
What is the core idea of such contraption?

[bson]

You need logic-level MOSFETs.  Such as Vishay Si4931DY, http://www.vishay.com/docs/72379/si4931dy.pdf .  I've used it to switch relays from 3.3V CMOS logic, even 3V latching telco relays worked fine off a 3.3V supply for me thanks to the low on resistance.

[sean87]

The OpAmp is in Non-Inverting mode, hence only 0V to 50V...The load is definitely is capable of drawing 4A. I need to switch between those resistors after output stage to control the circuit.

[sean87]

You need logic-level MOSFETs.  Such as Vishay Si4931DY, http://www.vishay.com/docs/72379/si4931dy.pdf .  I've used it to switch relays from 3.3V CMOS logic, even 3V latching telco relays worked fine off a 3.3V supply for me thanks to the low on resistance.

T

This has only 12V Vds...I need 50V/4A

[Zero999]

The OpAmp is in Non-Inverting mode, hence only 0V to 50V...The load is definitely is capable of drawing 4A. I need to switch between those resistors after output stage to control the circuit.

You've missed the point. What the op-amp is capable of driving is irrelevant. The maximum current is limited by the highest resistance part of the series circuit, which is R1 or R2, which you said are 1k and 10k. Ohm's law states that when the load is zero Ohms, the maximum current draw with the 1k resistor is 50/1000 = 50mA and 50/10000 = 5mA for the 10k resistor.

Unless you need a switch which bypasses R1 or R2, the maximum current the switch will ever see is 50mA, so 4A is overkill.

[T3sl4co1l]

The OpAmp is in Non-Inverting mode, hence only 0V to 50V...The load is definitely is capable of drawing 4A. I need to switch between those resistors after output stage to control the circuit.

Could you please address the numerous observations and questions the replies have raised.

Tim

[Hensingler]

http://catb.org/~esr/faqs/smart-questions.html

In particular

http://catb.org/~esr/faqs/smart-questions.html#goal